1. Technical Field
The present invention relates to a length measurement pattern for dimension measurement in forming a photoresist of a contact or a via on a diffusion layer or on an interconnect, as well as to a semiconductor device using this length measurement pattern and a method of manufacturing a semiconductor device.
2. Related Art
FIG. 3 is a view of a semiconductor device having a general multiple layer interconnect structure in which a contact 11 is disposed on a diffusion layer 10 which is the lowermost layer; an insulating film 12 is stacked thereon; a lower layer interconnect 13 and a via 14 are formed thereon; and an upper layer interconnect 15 is formed on an upper part of the via 14. In the most developed one of such semiconductor devices, the patterning dimension of the contact 11 and the via 14 is made superfine, so that it is necessary to measure the dimension of the patterned photoresist correctly so as to grasp the dimension and to perform centering with a good precision. When a management of this dimension is neglected, a dimensional abnormality or the like may possibly be generated that directly leads to decrease in the yield or to inconvenience of properties.
As a technique for correctly measuring the dimension of a photoresist patterned in this manner, a method using the length measurement SEM is known in the art (See Japanese Laid-Open patent publication NO. 2003-197503). In order to measure the photoresist dimension by the length measurement SEM, it is necessary to dispose a pattern for the length measurement SEM in a photoresist. In a photoresist dimension measurement pattern in the step of forming a via shown in FIG. 3, a contact pattern is disposed on a diffusion layer 10 which is the lowermost layer, and a via pattern is disposed on a lower layer interconnect 13, so as to perform measurement of the photoresist dimension by using the length measurement SEM. At this time, it is known that, with a pattern that has been made superfine to the same degree as the wavelength of the light used for exposure, a difference occurs in the photoresist dimension between an isolated pattern and a dense pattern by a physical phenomenon at the time of exposure.
FIG. 4 shows a conventional dimension measurement pattern of an isolated pattern.
Referring to FIG. 4, the dimension measurement pattern 16 which is an isolated pattern disposed in the lower layer interconnect 13 is disposed in an isolated state of only one, so that, naturally, no pattern is disposed in the neighborhood thereof. Therefore, at the time of dimension measurement of the photoresist with this dimension measurement pattern 16, the positioning and focusing in the length measurement SEM and the actual dimension measurement are carried out with the same dimension measurement pattern. Here, although not illustrated in the drawings, the photoresist dimension measurement of a contact 11 is carried out in the same manner with the dimension measurement pattern disposed in the diffusion layer 10.
The conventional photoresist has a resistance to electron beams. Therefore, even if the positioning and focusing of the length measurement SEM and the actual dimension measurement are carried out with the same dimension measurement pattern 16, no fatal fluctuation in the dimension or the like occurs, and no problem has been raised.
However, the photoresist used in an exposure apparatus using an ArF laser light source that is used in the currently most developed process and in the subsequent most developed lithography technique is weak in the resistance to electron beams. Therefore, it is known that, when electron beams are applied for a long period of time at the time of dimension measurement with the length measurement SEM apparatus used in the photoresist dimension measurement, the resist will be etched, making it impossible to carry out a correct dimension measurement.
With a currently used photoresist for an exposure apparatus using an ArF laser light source, a dimension fluctuation of about 1 nm/1 sec is seen in the photoresist dimension value when electron beams having an acceleration voltage of about 800 to 1600 V is radiated, though depending on the condition of the length measurement SEM. In the case of a photoresist dimension of a via, the dimension value is shifted to a larger side.
Since the newest via dimension is designed to be about several ten nm to a hundred and several ten nm, a large dimension measurement error occurs by the dimension fluctuation when an electron beam irradiation of several seconds to ten and several seconds is carried out in positioning and focusing.